s-news
[Top] [All Lists]

Re: Loess

To: Davorka Gulisija <gulisija@calshp.cals.wisc.edu>
Subject: Re: Loess
From: ripley@stats.ox.ac.uk
Date: Sun, 23 Mar 2003 19:59:13 +0000 (GMT)
Cc: s-news@wubios.wustl.edu
In-reply-to: <Pine.GHP.4.44.0303231342220.16592-100000@calshp.cals.wisc.edu>
You have just posted a slightly different version of this question on 
R-help.

- are you using S-PLUS or R?
- are you using loess.smooth or loess?  (note, no capitals!)

The differences matter somewhat.

On Sun, 23 Mar 2003, Davorka Gulisija wrote:

> I am using Loess procedure in order to infer certain relationship
> for the data set of ~130,000 observations with ~300 distinct values of
> single predictor. I understand that fitted values y-hat are just 300
> Weighted
> LS fits in certain neighborhood of predictors. I am bit confused about
> how exactly is this neighborhood assigned . Say I choose spanning
> parameter = .5, for each LS analysis 75000 observations should be used.
> However, intuitively it doesn't seem right since points are not equally
> distributed among predictors and there are many observations for a single
> value of predictor.
> I would appreciate if someone could clear this for me.

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595


<Prev in Thread] Current Thread [Next in Thread>